Sherif Abdelmohsen

General system architecture for BIM: An integrated approach for design and analysis

One of the significant benefits of Building Information Modeling (BIM) is the ability to effectively use analysis and evaluation programs during design, as feedback. However, the current dominant approach to analysis and evaluation of design proposals requires the creation of a separate building model for each kind of evaluation. This typically involves using a BIM tool to prepare the data for a specific type of analysis to obtain design feedback. Most of the effort lies in modifying the building model to support the analysis required. When dealing with multiple evaluations, this process is time consuming, greatly reducing the design benefits of BIM. We propose a system architecture to facilitate analysis and feedback in architectural design, based on post-processing design-oriented building models. The post-processing automatically adapts the building model to the needs of the specific analysis, where multiple analyses can be run from the same building model. We outline the methods for realizing such design interoperability. By utilizing geometric and attribute relationships and semantics, data subsets are identified and aggregated. We present an example where the design of a class of buildings - federal courthouses, is evaluated in terms of multiple analyses: programmatic spaces, building circulation, energy consumption, and preliminary cost. These analyses are performed by post-processing a single BIM model. The method is applicable to both API-based direct interfaces as well as open-standard building models.

An Inquiry into Designing in Context using Generative Systems

The use of generative systems has been widely investigated in the architectural design process through different procedures and levels of autonomy to generate form. The digression from abstract pre-existing notions of vocabulary and rules – even when resulting in emergent forms – to address complex real-world contexts is yet a challenging undertaking. This paper explores incorporating context in the process of designing using generative systems from ideation to fabrication, and explores the relationship between the emergent nature of generative design and the situated act of designing while using generative design tools. A course offered for 3rd year architecture students at the Department of Architecture, Ain Shams University, Egypt, was designed for this purpose. 110 students employed systems including shape grammars, L-systems, fractals and cellular automata, to design and fabricate 8 group projects. A discussion around emergence and situatedness is presented, with special attention to the designing process from ideation to fabrication.

A Heuristic Approach for the Automated Generation of Furniture Layout Schemes in Residential Spaces

A variety of heuristic methods and algorithms have been developed for space layout planning problems. Recent efforts to generate furniture layout schemes in existing spatial configurations have mostly relied on exhaustive search and are likely to produce dysfunctional or counter-intuitive solutions. In this paper, we propose a heuristic approach for the automated generation of furniture layout schemes, with specific focus on residential spaces. First, we present an operational definition for furniture entities, space configurations, and space entities. Then we introduce a heuristic algorithm for generating furniture layout schemes based on a set of space subdivision rules, object-object relations, and object-space relations. Using Grasshopper, we generate a group of possible schemes for a sample residential living space. A discussion follows, outlining current limitations, expanding the context of the study, and possibilities for development.

Rule-based vs. intuition-based generative design: an inquiry into the digital chain concept in undergraduate architectural education

Abstract While the digital chain concept has been explored in a number of architectural projects, only a few have pedagogically investigated the specific generative design methods in such a process. Different schools of thought have emerged in this regard. Some promote the rule-based approach, where specific rules are more dominant than inspiration and intuition, and vice versa. Based on an inquiry into the digital chain concept within the context of a digital design course, we argue that there exists a set of dynamics that govern the implemented generative methods, that is neither fully rule-based nor intuition-based, but a hybrid of both. Revisiting the digital chain based on a hybrid intuition-based/rule-based process: (a) concept generation based on intuition and inspirations, (b) back and forth digital experimentation and material exploration, (c) informed rule-based and algorithmic translation based on feedback from conceptual departure, intuition, and digital experimentation, (d) parametric modeling process, (e) preparation for fabrication based on material exploration and parametric modeling

Optimized design to robotic production: integrating multipoint forming and free form material deposition in the mass customization of double curved facades

Abstract The mass customization of double curved façade panels is still in its infancy, in spite of recent efforts in panel manufacturing techniques. As these techniques have typically developed workarounds for curved façade manufacturing, they present challenges regarding economic feasibility, façade surface continuity, accuracy and waste reduction. We present an optimized design-to-robotic production approach that integrates multipoint forming and adjustable moulds together with robotic free form material deposition to produce mass customizable double curved façade panels. We argue that our approach generates customizable panels with complex geometry, and optimizes fabrication workflow, with minimum waster and maximum accuracy in representing irregular forms. Overview of the proposed optimized design-to-robotic production workflow: (a) design of panel to be fabricated, (b) flexible mechanical modular system and silicone sheet to interpolate surface curvature from underlying modules, (c) tool paths generated from the digital model normal to the underlying base surface, (d) robotic arm deposits material over the modular system, (e) low polygon preview of the material deposition based on the tool paths of the robotic arm.